[UA-discuss] Regarding "RTL"

Mark Svancarek marksv at microsoft.com
Tue Feb 24 22:56:05 UTC 2015


Yes, this is my concern.  I think we have an opportunity to set well-defined expectations and ensure implementations converge to them.

Fully right-to-left email addresses (with a RTL local-part) are pretty rare, so while there may be precedent, I don't think the behavior is well-defined.


From: ua-discuss-bounces at icann.org [mailto:ua-discuss-bounces at icann.org] On Behalf Of Brent London
Sent: Tuesday, February 24, 2015 2:52 PM
To: Dusan Stojicevic
Cc: ua-discuss at icann.org
Subject: Re: [UA-discuss] Regarding "RTL"

1) Where are e-mail solution providers gathering to discuss this?   Does a forum already exist?  if not, should the UASG create one?

I think the appropriate forum for this would be m3aawg<https://www.maawg.org/>, although I'm unsure of whether it's actually been raised there yet.

2) Brent:  How has Gmail approached this?  Ignored the dots in the username side of the address?

Gmail uses banner notifications to warn users when there's something problematic about a message, e.g., a suspicious From heade! r. See the first expandable section of this<https://support.google.com/mail/answer/1366858?hl=en> article for more info.

Gmail ignores dots in the usernames of gmail.com<http://gmail.com> addresses, which is doable because Gmail controls the gmail.com<http://gmail.com> namespace. So hello.world at gmail.com<mailto:hello.world at gmail.com> goes to the same user as helloworld at gmail.com<mailto:helloworld at gmail.com>. We can't, however, do that for other domains. For example, helloworld at outlook.com<mailto:helloworld at outlook.com> and hello.world at outlook.com<mailto:hello.world at outlook.com> are different users.

So, consumer mail clients are prepared with the LTR rule: that on the left side from "@" sign is name, and on the right side is domain name.
But, they don't check this rule.
I presume that in the RTL world, consumer mail clients do the same - with RTL rule.

I think the circumstances are different in the RTL world. Fully right-to-left email addresses (with a RTL local-part) are pretty rare, so while there may be precedent, I don't think the behavior is well-defined.












Brent London
brentlondon at google.com<mailto:brentlondon at google.com>
+1 650-214-5206

On Tue, Feb 24, 2015 at 12:16 AM, Dusan Stojicevic <dusan at dukes.in.rs<mailto:dusan at dukes.in.rs>> wrote:
Dear all,

We have RTL problem in LTR world already. Try this.
F.E. / I will use ASCII just to make a point>
com.gmail at stojicevic.dusan<mailto:com.gmail at stojicevic.dusan>
I send one email to this mail address and it gone "as usual".
After a minute, I've got a massage in attach.

So, consumer mail clients are prepared with the LTR rule: that on the left side from "@" sign is name, and on the right side is domain name.
But, they don't check this rule.
I presume that in the RTL world, consumer mail clients do the same - with RTL rule.
The real question is - should mail clients check this?

Regards,
Dušan

On 24.2.2015 2:17, Mark Svancarek wrote:
I’ve seen some discussion activity at w3.org<http://w3.org> (as recently as last December https://www.w3.org/International/wiki/EAI_Address_Issues ) but it doesn’t seem actionable.  I think we’ll need to use this group to get all similar issues clarified and make them usable by developers.

From: ua-discuss-bounces at icann.org<mailto:ua-discuss-bounces at icann.org> [mailto:ua-discuss-bounces at icann.org] On Behalf Of Don Hollander
Sent: Monday, February 23, 2015 4:46 PM
To: Brent London
Cc: ua-discuss at icann.org<mailto:ua-discuss at icann.org>; Edmon Chung
Subject: Re: [UA-discuss] Regarding "RTL"

Thanks Brent, Alireza & Edmon.

This is a very interesting question of the approach that a software supplier might take.

I have two questions:

1) Where are e-mail solution providers gathering to discuss this?   Does a forum already exist?  if not, should the UASG create one?

2) Brent:  How has Gmail approached this?  Ignored the dots in the username side of the address?

I think it’s very interesting.

Don


On 24/02/2015, at 1:18 pm, Brent London <brentlondon at google.com<mailto:brentlondon at google.com>> wrote:

It becomes problematic, as Edmon mentioned, when there are dots in both sides. It's especially confusing if both sides contain a string that plausibly could be a TLD:

customer.care@شزذ.يثب<mailto:customer.care@%D8%B4%D8%B2%D8%B0.%D9%8A%D8%AB%D8%A8>











Brent London
brentlondon at google.com<mailto:brentlondon at google.com>
+1 650-214-5206<tel:650-214-5206>

On Sun, Feb 22, 2015 at 5:46 AM, Alireza Saleh <saleh+ua at nic.ir<mailto:saleh+ua at nic.ir>> wrote:
There is no problem as long as the usernames starts and ends with a character with a property of AL (Arabic Letter), R (Right to left) or L (Left to right) otherwise in a LTR context it may jump around @ sign and make the address unreadable. The email address may also become unreadable in LTR context If the username part starts with L and ends with AL like :

testمثال@مثال.تست<mailto:test%D9%85%D8%AB%D8%A7%D9%84@%D9%85%D8%AB%D8%A7%D9%84.%D8%AA%D8%B3%D8%AA>

the red part is username.

-Alireza


On Feb 22, 2015, at 11! :47 AM, Edmon Chung <edmon at registry.asia<mailto:edmon at registry.asia>> wrote:

What about where the username part contains a dot or other separators? Is there a difference between “.” And “-“ or “_”?

tld.sld at name.user<mailto:tld.sld at name.user> ?
tld.sld at name-user<mailto:tld.sld at name-user> / tld.sld at user-name<mailto:tld.sld at user-name> ?
etc.?

Edmon




From: ua-discuss-bounces at icann.org<mailto:ua-discuss-bounces at icann.org> [mailto:ua-discuss-bounces at icann.org] On Behalf Of Alireza Saleh
Sent: Sunday, February 22, 2015 4:07 PM
To: Edmon Chung
Cc: ua-discuss at icann.org<mailto:ua-discuss at icann.org>
Subject: Re: [UA-discuss] Regarding "RTL"

This is very interesting question. I’ve also thought about it before. This is a new topic and there is no similar experiences. I don’t have exact answer to this question but overall I think people mainly prefer the RTL version with right alignement. however the bidi property of @ allows its usage in the middle of RTL texts without creating any confusions unlike <http://>.

نام@مثال.آزمایشی<mailto:%D9%86%D8%A7%D9%85 at xn--mgbh0fb.xn--hgbk6aj7f53bba>
TLD.SLD at NAME<mailto:TLD.SLD at NAME>


-Alireza



On Feb 22, 2015, at 4:01 AM, Edmon Chung <edmon at registry.asia<mailto:edmon at registry.asia>> wrote:

That applies to email (EAI) addresses as well I suppose?
Which Brent has been bringing up.
So, within a RTL ! context (e.g. if the user interface or other elements are RTL) one should expect

tld.domain at name.user:mailto<mailto:tld.domain at name.user:mailto>

Is that correct Alireza?

Edmon



From: ua-discuss-bounces at icann.org<mailto:ua-discuss-bounces at icann.org> [mailto:ua-discuss-bounces at icann.org] On Behalf Of Alireza Saleh
Sent: Sunday, February 22, 2015 2:54 AM
To: Mark Svancarek
Cc: ua-discuss at icann.org<mailto:ua-discuss at icann.org>
Subject: Re: [UA-discuss] Regarding "RTL"

Dear Mark,

Just a quick note about your question, it is expected the label starts from the right side of the address baar, and from right to left. So the main issue would be the alignment. Natively it should look like:

com.microsoft.www://http

-Alireza



On Feb 20, 2015, at 10:35 PM, Mark Svancarek <marksv at microsoft.com<mailto:marksv at microsoft.com>> wrote:

Hi, I had some questions regarding my recent usage of the term “RTL”.  By this I mean “right to left”, a characteristic of Arabic and Hebrew.  At Microsoft we also call this “bidi” (bidirectional).

Here’s a discussion regarding RTL.  (I’ve also attached a much more detailed explanation, which includes Microsoft’s recommendations, but it’s in PowerPoint.  Hopefully you already! use a compatible viewer.)
Bidi display of IRIs (URLs/URIs)
Bidirectional display of IRIs (an IRI with some Right-To-Left characters, eg: Arabic) has some odd quirks.  There’s an IETF WG working on creating an IRI RFC.  It’d be nice if we could help ensure that there were reasonable standards for the display of bidi IRIs.  The existing IRI drafts suggest using the Unicode Bidi Algorithm to display IRIs, but that has some problems.

User and government feedback indicates that our current behavior is a bit unexpected.  Currently we have some odd quirks about the display of Bidi IRIs in Microsoft.  This is just an example, other places may have different odd quirks.

Logical Order

IE with LTR context

IE with RTL context

http://www.microsoft.com<http://www.microsoft.com/>

http://www.microsoft.com<http://www.microsoft.com/>

http://www.microsoft.com<http://www.microsoft.com/>

http://اا1اا.بب2بب.ةة3ةة<http://xn--1-ymcaba.xn--2-0mcaba.xn--3-2mcaba/>

http://اا1اا.بب2بب.ةة3ةة<http://xn--1-ymcaba.xn--2-0mcaba.xn--3-2mcaba/>

http://اا1اا.بب2بب.ةة3ةة<http://xn--1-ymcaba.xn--2-0mcaba.xn--3-2mcaba/>

http://a1a.اا2اا.بب3بب.d4d<http://a1a.xn--2-ymcaba.xn--3-0mcaba.d4d/>

http://a1a.اا2اا.بب3بب.d4d<http://a1a.xn--2-ymcaba.xn--3-0mcaba.d4d/>

http://a1a.اا2اا.بب3بب.d4d<http://a1a.xn--2-ymcaba.xn--3-0mcaba.d4d/>

http://اا1اا.b2b.c3c.بب4بب<http://xn--1-ymcaba.b2b.c3c.xn--4-0mcaba/>

http://اا1اا.b2b.c3c.بب4بب<http://xn--1-ymcaba.b2b.c3c.xn--4-0mcaba/>

http://اا1اا.b2b.c3c.بب4بب<http://xn--1-ymcaba.b2b.c3c.xn--4-0mcaba/>


...

[Message clipped]

---------- Forwarded message ----------
From: Mail Delivery Subsystem <mailer-daemon at googlemail.com<mailto:mailer-daemon at googlemail.com>>
To: dusan at dukes.in.rs<mailto:dusan at dukes.in.rs>
Cc:
Date: Tue, 24 Feb 2015 08:04:19 +0000
Subject: Delivery Status Notification (Failure)
Delivery to the following recipient failed permanently:

     com.gmail at dusan.stojicevic<mailto:com.gmail at dusan.stojicevic>

Technical details of permanent failure:
DNS Error: Address resolution of dusan.stojicevic. failed: Domain name not found

----- Original message -----

X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net<http://1e100.net>; s=20130820;
        h=x-gm-message-state:message-id:date:from:user-agent:mime-version:to
         :subject:content-type:content-transfer-encoding;
        bh=oV7WQ8lryXPSsTVyIzYLp3QZdFh+3U95bpv4Okv/xjI=;
        b=QqLxV2xWblSmXtEvb8ak92Y3nnpQMC6pqDr5pp817ubddgmuUJqPdviXYpj0UTqa+V
         X8+F7jISs2TUpugEXyFYwG9OVo9AbD5tbgvYd+L+sZhEicjmJ1VkK10yCcj2g/4Lll1F
         NQjLy+uL7xhiXudx3DRVN3FebiZ9MT2mf8NWNHfJzodeTiQqfTZ39spVDzuZJUoID3WR
         j4wSf1iVpBKximpwxU/PXSPF1exepSDJBGNSk/XX6DCKEpl3AmemheU/52Bs1xO5Cf/i
         uEt66tW5W5ufTLKtRaOH/BnPFHTUtpnT26txciwpZI8zuGPm0GNefFFhbOwQjggFxLrf
         93hA==
X-Gm-Message-State: ALoCoQlQhBKx8PrtsJjuCaKmlNYZK/+74aUgVNjwV2wkNFRmTp/ZveDMMVEMS/l7lo4b9aGuvha/
X-Received: by 10.180.77.166 with SMTP id t6mr28567023wiw.28.1424765057763;
        Tue, 24 Feb 2015 00:04:17 -0800 (PST)
Return-Path: <dusan at dukes.in.rs<mailto:dusan at dukes.in.rs>>
Received: from [127.0.0.1] (178-221-219-238.dynamic.isp.telekom.rs<http://178-221-219-238.dynamic.isp.telekom.rs>. [178.221.219.238])
        by mx.google.com<http://mx.google.com> with ESMTPSA id s5sm23757985wia.1.2015.02.24.00.04.15
        for <com.gmail at dusan.stojicevic<mailto:com.gmail at dusan.stojicevic>>
        (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128);
        Tue, 24 Feb 2015 00:04:16 -0800 (PST)
Message-ID: <54EC3080.1080807 at dukes.in.rs<mailto:54EC3080.1080807 at dukes.in.rs>>
Date: Tue, 24 Feb 2015 09:04:16 +0100
From: Dusan Stojicevic <dusan at dukes.in.rs<mailto:dusan at dukes.in.rs>>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0
MIME-Version: 1.0
To: com.gmail at dusan.stojicevic<mailto:com.gmail at dusan.stojicevic>
Subject: proba
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: quoted-printable
X-Antivirus: avast! (VPS 150223-1, 02/23/2015), Outbound message
X-Antivirus-Status: Clean



---
Ova e-pošta je provjerena na viruse Avast protuvirusnim programom.
http://www.avast.com


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mm.icann.org/mailman/private/ua-discuss/attachments/20150224/f706673b/attachment.html>


More information about the UA-discuss mailing list